Proteomic analysis of ascitic extracellular vesicles describes tumour microenvironment and predicts patient survival in ovarian cancer

Abstract

High-grade serous carcinoma of the ovary, fallopian tube and peritoneum (HGSC), the most common type of ovarian cancer, ranks among the deadliest malignancies. Many HGSC patients have excess fluid in the peritoneum called ascites. Ascites is a tumour microenvironment (TME) containing various cells, proteins and extracellular vesicles (EVs). We isolated EVs from patients' ascites by orthogonal methods and analyzed them by mass spectrometry. We identified not only a set of 'core ascitic EV-associated proteins' but also defined their subset unique to HGSC ascites. Using single-cell RNA sequencing data, we mapped the origin of HGSC-specific EVs to different types of cells present in ascites. Surprisingly, EVs did not come predominantly from tumour cells but from non-malignant cell types such as macrophages and fibroblasts. Flow cytometry of ascitic cells in combination with analysis of EV protein composition in matched samples showed that analysis of cell type-specific EV markers in HGSC has more substantial prognostic potential than analysis of ascitic cells. To conclude, we provide evidence that proteomic analysis of EVs can define the cellular composition of HGSC TME. This finding opens numerous avenues both for a better understanding of EV's role in tumour promotion/prevention and for improved HGSC diagnostics.

Keywords: ascites; extracellular vesicles (EV); fallopian tube and peritoneum (HGSC); high-grade serous carcinoma of the ovary; macrophage; ovarian cancer (OC); tandem mass spectrometry (MS/MS); tumour microenvironment (TME).

FIGURE 1

Isolation and characterization of EVs. (a) Schematic overview of processing and isolation of samples from patient ascites prior the mass spectrometry (MS) analysis. Three different samples were isolated from each patient ascites. Samples U stand for EVs isolated by differential ultracentrifugation coupled with sucrose cushion purification step; samples S and B were both obtained from size‐exclusion chromatography of concentrated ascitic fluid: S fraction contains the majority of EVs, whereas B fraction contains the bulk of proteins and served as background of contaminating proteins for each patient sample. Due to principle of each method, different input volumes of ascites were used: 38 mL for sample U and 2.5 mL (concentrated to 1 mL) for samples S and B. (b) Representative Cryo‐EM images of samples—close‐up in the top and overview in the bottom. Red arrows in S images show EVs present in the overview image (due to their low count). (c) DLS measurements of EVs. Each graph shows average of three separate measurements for the sample. Size and polydispersity index are reported as mean ± standard deviation (SD), concentration as mean (n = 3). (d) WB analysis for EVs markers/EV‐associated proteins and a ‘negative marker’/non‐EV associated protein ApoA‐I. (e) Amidoblack staining for total protein on the membrane. Staining was performed after WB analysis.

FIGURE 2

Characterization of isolated EV isolates by tandem mass spectrometry. (a) Barplot showing number of proteins identified by mass spectrometry in each fraction for individual patients. (b) The Pearson's correlation of log transformed data of all protein intensities detected by MS in EVs isolated by UC and SEC in patients 1–11. Blue line represents the linear regression. (c) Heatmap depicting overlay of mass spectrometry data with the list of EV‐associated proteins (five categories) published in MISEV2018 (Théry et al., 2018). U = EVs isolated by ultracentrifugation, S = EVs isolated by size‐exclusion chromatography, B = bulk of non‐EV proteins, isolated by size exclusion chromatography.

FIGURE 3

Identification and characterization of core ascitic EV‐associated proteins. (a) Schematic overview of processing of data obtained from mass spectrometry analysis. Data were processed in a patient‐dependent manner. S = proteins from EVs isolated by size‐exclusion chromatography, U = proteins from EVs isolated by ultracentrifugation, S&U = protein is present in both S and U, B = bulk of non‐EV proteins isolated by size exclusion chromatography, that are subtracted from S&U. Only data from S&U (lime green) are further analyzed through this figure. (b) UpSet plot visualizing intersections of sets of proteins between patients. Only sets present in ≥9 patients are shown (S&U‐80). One hundred one of these proteins are present in all 11 patients (S&U‐100). (c) Gene ontology analysis for the top 10 cellular compartments of S&U‐100 and S&U‐80 identified by both methods. FDR—false discovery rate. (d) Stacked barplots showing cellular localization for S&U‐100 and S&U‐80. The analysis is based on results from Human Cell Map project (Go et al., 2021). NMF, non‐negative matrix factorization. Approximately half of the proteins from both groups were not identified in this database.

FIGURE 4

Identification of ‘HGSC‐specific EV‐associated proteins’. S = proteins from EVs isolated by size‐exclusion chromatography, U = proteins from EVs isolated by ultracentrifugation, S&U = protein is present in both S and U. (a) Schematic overview of identification of eight HGSC‐specific EV‐associated proteins. (b) Presence of eight HGSC‐specific EV‐associated proteins among patient samples. (c) Western blotting for selected HGSC‐specific EV‐associated proteins on U fractions of 11 patients and Caov‐3 and Kuramochi HGSC cell lines. Caov‐3 and Kuramochi cells are used as negative controls for macrophage markers MRC1, CD68, and FCGR1A. For TACSTD2, Caov‐3 serves as a positive control, while Kuramochi serves as a negative control. (d) Schematic overview of identification of 157 HGSC‐specific EV‐associated proteins. These data were augmented by proteins identified by UC.

FIGURE 5

Cells of origin of EVs. (a) Scheme depicting various origins of EVs in ascites. (b, c) Heatmaps visualizing expression of genes of interest. Data and clusters are based on analysis by (Izar et al., 2020). (Gene names in heatmap C are legible upon zoom in).

FIGURE 6

Analyses of ascitic EVs by cell of origin. (a) Heatmap visualizing presence of the cell type‐specific markers, by (Izar et al., 2020) found in our mass spectrometry data of EVs samples isolated by UC (a) and SEC (a’). (b) Graph showing proportions of cell types according to protein intensities of cell type‐specific markers for each patient. Data are based on protein intensities measured on EVs samples isolated by UC (b) and SEC (b’). (b’’) shows correlation of the (b) and (b’) data. CI, confidence interval.

FIGURE 7

Analyses of ascitic cells. (a) Scheme of the analysis of ascitic cells and EVs. (b) Table of markers used for FC. (c) TSNE plot of markers detected by spectral flow cytometry on all ascitic cells. (d) Graph showing proportions of cell types detected by FC in ascites of individual patients using FlowSOM clustering. (d’) shows proportions of malignant cells, fibroblasts and macrophages detected by FC in ascites of individual patients after removal of populations of peripheral blood cells (possible contaminants). (e) The repeated measure correlation between proportions of major cell populations derived from FC data and from measurements of EV proteins. Data are based on protein intensities measured on EVs samples isolated by UC (e) and SEC (e’). CI—confidence interval. (f) Graphs showing ‘enrichment’ of EVs versus cells in individual ascites. The enrichment coefficient (EC) is defined as fold enrichment of (% EV)/(% cells). Proportions of cell types based on protein intensities measured on EVs samples isolated by UC (f) and SEC (f’) were plotted against proportions of major cell types (according to Izar et al., 2020) detected by FC.

FIGURE 8

Survival analyses. (a) Scheme of the analysis which cellular and EV parameters in ascites correlate with overall survival of HGSC patients in our study. (b) Overall survival of patients based on the composition of ascitic EVs. Log‐rank (Mantel‐Cox) Test on protein intensities in U samples. (c) Forest plot of OS analysis in b). HR, hazard ratio. Extended data: UpSet plot visualizing all intersections of sets of proteins identified in EVs isolated by both methods. Each of 2418 proteins in this graph was recovered after both SEC and UC simultaneously in at least one patient. Intersections highlighted in lime green consist of sets of proteins found in ≥9 patients.